EFFECTS of KNIFE JOINTING and WEAR on the PLANED SURFACE QUALITY of NORTHERN RED OAK WOOD Roger E

Home , Jointing (sharpening)

EFFECTS OF KNIFE JOINTING AND WEAR ON THE PLANED SURFACE QUALITY OF NORTHERN RED OAK WOOD Roger E. Hernu'ndez' Professor

and Luiz Fernando de Moura Graduate Student DCpartemcnt des Sciences du Bois et de la For&t Universite Laval QuCbec, Canada GI K 7P4 (Received November 2001)

ABSTRACT

Jointing is a technique to obtain the same cutting circle for all knives mounted in a cutterhead of a peripheral knife planer. Initially the jointed land at thc cutting edge has a 0 degree clearance angle that becomes negative with workpiecc motion relative to the cutterhead and as the cutting edge wears. Jointed knives may crush cells on the planed surface and affect the quality and performance of wood for end ~lscs.Wc evaluated the gluing properties of northern red oak planed surlhceh that had been planed using one of three jointed land widths, over four levels of knife wear. Under these cutting conditions. surface roughness significantly influenced gluability more than cellular damage. In sum, gluing performance was positively affected by knife wear, and no variation in gluing performance iumong the jointcd land widths studied existed. In samples after accelerated aging, the effects of wear on gluing were more pronounced, with an improvement in gluing performance, associated with an increaw in surface roughness and permeability with increased knife wear. These results suggest a jointed land of 1.2 tnm as the maximum allowable width for planing red oak wood prior to gluing. Also. thc planed surfacc gluability of this wood may be enhanced using a knife with the rake face recession of 132 pm and the clearance face recession of 438 pm, which results in a surface roughness ol' 10 prn R ,,,,,,.

Kr\.n,o~.rl\: Planing. knife iointing. wear. gluing properties, northern red oak

IN~I~IIOIIIJC'I'IONAN11 BACKGROIJND The grinding action forms a jointed land on Jointing is a common practice applied to pe- the cutting edges where width varies accord- ripheral knife planers to produce an equal cut- ing to the initial knife projection from the cutting circle for all knives mounted in a cutter- terhead with respect to the other knives. The head. An abrasive stone is passed over the nominal knife clearance angle returns to 0 de- knife edges as the cutterhead turns at its nor- grees in such a manner that during planing the mal cutting speed. Any projecting knife edge face of the land causes perpendicular com- is ground back. ensuring that all the edges lie pression as well as friction on the wood sur- in a common cutting circle. Thus, each knife face. These forces normally should increase can work in a uniform manner taking chips of with the increased jointing. The feeding action equal thickness (Dunsmore 1965; Hoadley and wear at the cutting edge may also cause 2000). Jointing makes knife edges more wear- the clearance angle to become negative during resistant, and it is sometimes wrongly repeated cutting, reduce the rake angle, and further in- as a sharpening process. crease the normal component of the cutting forces. As a result, surfaces planed with joint- ' Mcmbcr ol SWST. ed knives are subjected to greater crushing and II~~.~!ii~zilc;(III~(I(' MoI(Io-E~F.ECTS 01: JOINTING AND WEAR ON PLANED SURFACE QUALITY OF RED OAK 54 1 friction than surtices machined with unjoined oak, there was no apparent cell damage and knives. Although greater cutting forces are in- gluing performance was improved as jointed volved with jointing, a positive or negative ef- land width increased. This was associated with fect on gluing may result, depending on the an increase in the permeability of the wood. species of wood (Hernandez and Naderi However, planing tests in this earlier study 200 I). were carried out with freshly sharpened knives Damage on the surface and subsurface of only. Hernandez and Naderi (200 1 ) concluded wood plays an important role on the gluing that information was needed on interactions performance (River and Miniutti 1975; Jokerst among anatomical wood structure, jointed et Stewart 1976; White and Green 1980; Mur- land width, and knife wear. manis et al. 1983; Caster et al. 1985; Hernan- In addition to jointed land width, acceptable dez 1994; Hernandez and Naderi 2001; Her- limits of knife wear for planing wood need to nlindez and Rojas 2002). The type and level be established. However, few data are avail- of damage can affect subsequent transforma- able on the interaction between jointing and tion processes or even the performance of the wear. In general, differences in gluing prop- final product (Stewart and Crist 1982). A erties have been reported among wood species poorer glueline performance has been associ- (Hernandez and Naderi 200 1) and data are re- ated with the presence of damaged cell layers quired for different wood species. on the glued wood surface (River and Miniutti This work evaluated the planed surface 1975; Jokerst and Stewart 1976). In general, quality of northern red oak wood as a function crushed cells may obstruct adhesive penetra- of jointed land width and knife wear. Four tion (Caster et al. 1985). Superficial crushed wear levels of three jointed land widths were cells are a result of high cutting forces gen- tested. The gluing shear performance was erated by knives during the planing operation evaluated at constant hygrothermal conditions (Jokerst and Stewart 1976). Hernandez and and after exposing surfaces to an accelerated Naderi (2001) reported that the depth of the aging treatment. crushed cell layer increased as the width of the jointed land of knives increased. Therefore, MATERIALS AND METHODS there should be a maximum jointed land width with this jointing technique where the cutting Testing materials efforts continue to result in a good quality sur- Northern red oak (Qu~rcusruhrrr L.) wood, face. a ring-porous hardwood, commonly applied Hoadley (2000) suggested that a very nar- indoors was tested. Prior to testing, the com- row land (0.25 mm) can be used without any mercial air-dried lumber was stored in a con- loss in quality because the area of zero clear- ditioning room at 60% relative humidity (RH) ance is very small. Koch (1985) and the and 20°C for 6 months. At the end of this pe- ASTM D 1666 (1987) indicate that the max- riod, the average moisture content (MC) was imum jointed land width before re-sharpening 10%. Eighty boards, 63 mm wide, 20 mm should be 0.8 mm. Dunsmore (1965) recom- thick, and 2,220 mm long, were selected and mended a maximum of three or four jointings, prepared for planing tests. The boards were with a final jointed land width up to 1.2 mm. separated into four groups of twenty, with However, Hernandez and Naderi (2001) sug- each group to be subjected to a specific wear gested that the maximum allowable width for level. Each board was cross-cut into three the jointing operation depends on the wood matched sections 550 mm long, and these species. They found that sugar maple had per- were assigned for testing to a specific jointing manent damage to the wood surface and con- level. This meant that a total of twelve subsequently, the gluing performance decreased groups of twenty specimens each were tested. as the jointed land width increased. For red Each section underwent a planing treatment 542 woo11 AND FIBER SCIENCE. OCTOBER 2002. V. 34(4) and was then sectioned to prepare a laminated mensions of the laminated blocks were 50 mm block and a specimen for measuring cutting wide, 38 mm thick, and 235 mm long. Two forces. gluing shear samples were prepared from each laminated block. A total of 40 samples from Specimen yrepcrrurion each group was selected to evaluate the effect of knife jointing and wear on the gluing Each group was surfaced with a specific strength after exposure to two hygrothermal jointed land knife and at the assigned wear conditions. The gluing shear test samples were levels with a Weinig Unimat 23EL moulder. machined according to ASTM D 905 (1994). The feed rate was set to give 20 knife marks per 25 mm of length and the cutting depth was Conditioning treatments adjusted to remove 1 mm of wood in one pass. The cutting circle radius was 75 mm. One of The samples were prepared under constant the three knives on the cutterhead was set for conditions of 60% RH and 20°C, and half cutting. The knife angle and the clearance an- were tested under these conditions. The other gle for the freshly sharpened knives were 45 half were subjected to accelerated aging, and 25 degrees, respectively. After jointing, where samples were immersed in distilled wa- these angles became 70 degrees for the knife ter for 6 h, resulting in an average MC of angle and 0 degrees for the clearance angle. about 27%. They were then oven-dried at The rake angle was 20 degrees. The sharp 40°C for 18 h. This soaking-and-drying cycle knives were jointed with a silicon carbidelalu- was repeated twice. Finally, all samples were minum oxide abrasive stone at two specific conditioned at 60% RH and 20°C to reach jointed land widths, 0.4 and 1.2 mm. A third their initial 10% equilibrium MC. group of boards was planed with an unjointed knife, and was considered as having a 0-mm Glueline tests jointed land width. One knife was successively Samples were evaluated according to the worked up to reach four wear levels for each ASTM D 905 standard (1994) with a universal of the three jointing levels. Thus, the process testing machine fitted-out with a gluing shear began with a planing test by a freshly sharp- fixture. The crosshead speed was 5 mmlmin. ened knife followed by other planing tests af- Cross sections of the samples and load at fail- ter 500 m, 1,000 m, and 3,000 m of planing. ure were recorded, and the average gluing To generate wear, two-hundred ninety knot- shear stress was calculated. The percent wood free pieces of northern red oak wood 75 mm failure was then independently estimated by wide by 25 mm thick and of varying lengths two technicians. The statistical analysis of the were planed under the same cutting conditions results was performed using multivariate and as during the planing treatment itself. After univariate analysis of variance. this planing treatment, the boards were cross- cut to form pairs that were subsequently glued Microscopic evaluation together to form a laminated block. Small blocks with about 0.8 cm2 of trans- The pairs of boards were glued using a car- verse area, including the glueline, were re- pentry white adhesive (polyvinyl acetate), moved from the laminated block for scanning with a glue consumption of 8 10 g per square electron microscopy (SEM). A surface on the meter. A 0.4 MPa clamping pressure was ap- end-grain perpendicular to the planed surface plied for 45 rnin as recommended by the ad- was carefully cut with a razor blade. The hesive manufacturer. blocks were then desiccated with phosphorous Finally, the laminated blocks were ma- pentoxide over two weeks, mounted onto stan- chined to correct for any misalignment from dard aluminium stubs with silver paint, and sliding during the gluing process. The final di- coated with gold-palladium in a sputter-coater. Herrl~~rrtlc:trnd c/o Mour(r-EFFECTS OF JOINTING AND WEAR ON PLANED SURFACE QUALITY OF RED OAK 543

TABL~1. Sltrntncrn of' the ctlge r-ec.ession and the normtrl (Fn) and parallel (Fp) orthogonal cutting forc,e.sfrom,fi)ur plarlitzg lrtl~rhsutzd three jointed land widths.

Hdgc recr\\~,rn' Cutt~ngforces' In Jwntcd I,md lurrgth (~rr) (n~n?) Il:ikc face (&m) Clcar;incr I;lcr. (pn) F,, (nlmm) PI>(nlmni)

Electron micrographs of four representative and parallel components of the orthogonal cut- subsurfaces were taken for each of the twelve ting forces, 90 "-0 ", were measured. Knives cutting conditions. The depth of damage and were removed from the planer and mounted glueline thickness at five systematically cho- on the column of a milling machine. A cross- sen points from each SEM micrograph were ring dynamometer (King and Foschi 1969) evaluated. Surface roughness was estimated was fixed to the feed table to measure the cut- from the micrographs from the Rmax index. ting forces. This dynamometer was calibrated by known forces applied to the normal and Warmeasurements parallel directions. The feed speed was 280 A resin molding technique was applied to mmlmin with a chip thickness of 1 mm. The make replicas of the three type M-2 high nominal rake angle was 20 degrees; however, speed steel knives from the experiments. Sev- the knife geometry would have varied accord- en equally spaced cross-sections were taken ing to the wear level. During each test, the along the cutting edge area from each replica. cutting forces were recorded with a computer Micrographs of these cross-sections were tak- and a data acquisition card, set at 16 readings en with a digital camera mounted on the mi- per second. The minimum, average, and max- croscope and processed with an Adobe Pho- imum cutting forces for each test were deter- toshop 4 software and a Regent Instruments mined from these data. Wince11 Pro 5.6d image analyzer. The edge recession from an ideally sharp edge was mea- RESU1,TS AND DISCUSSION sured parallel to the rake and clearance faces Weur of knives. The quality of sharpening prior to the planing treatment was not uniform for the freshly Cutting ,force meusurements sharpened knives (Table 1). For this reason, Primarily radial oriented 7-mm-wide by 19- wear values after planing treatments were ad- mm-thick and 75-mm-long samples were tak- justed by subtracting observed wear at 0 m of en from the matched sections to evaluate the planing. Knife wear on the rake and clearance effect of jointing and wear on the cutting forc- faces increased as the jointed land width and es. After each planing treatment, the normal the planing length increased. The edge reces- 544 WOO11 AND FIBER SCIENCE. OCTOBER 2002, V. 34(4)

the friction force and the temperature would increase, increasing wear. This phenomenon could also occur between the rake face and the wood chips, being influenced principally by the rake angle and the chip thickness (Stewart 1991). In this study, chip thickness remained constant and the rake angle probably decreased given that recession on the rake face 1.2 occurred. The high temperatures during cutting could also have favored chemical attack & .o V. by extractives (tannic acid) increasing knife Y wear (Krilov 1986; Stewart 1989).

tiI. Edge recession over the knife clearance face Cutting as a function of the planing length for three levels of joint- forces ed land width (J.L.). Cutting forces were measured under conditions of experimental orthogonal slow-cut- sion was similar on both rake and clearance ling rather than under practical peripheral cut- faces of the knives. The adjusted edge reces- ting. Although experiments at low cutting sions on the clearance face as a function of speeds could be done at practical working planing length and jointed land width are giv- speeds (McKenzie 1960), it would not be pos- en in Fig. 1. In general, the increase in wear sible to analyze the negative clearance angle was high between 0 and 1,000 m of planing. due to the feeding action during peripheral Afterwards, wear increased at a slower rate up planing. The variation in the average orthog- to 3.000 In of planing. The wear pattern on onal cutting forces of northern red oak was the rake face was similar to that on the clear- greater than that reported by HernBndez and ance face. Kirbach and Bonac (1982) also re- Rojas (2002) for sugar maple wood tested un- ported that wear characteristically follows two der similar conditions. This variation was also stages, an initial rapid and progressive stage greater for the normal component (average followed by a second slower stage wear that C.V. 22%) compared to the parallel compo- tends to level off. We found a similar trend in nent (average C.V. 17%) of cutting forces (Ta- the wear process. ble 1). The higher variation for red oak wood The wear could be attributed principally to could be attributed to its ring-porous structure. simultaneous chemical and mechanical actions The minimum and maximum values of cutting (Hillis and McKenzie 1 964; Krilov 1986). The forces (not shown) were also highly variable friction generated between the knife rake face and generally associated with earlywood and and the chips produced and between the clear- latewood regions, respectively. ance face and the new wood surface could The cutting force in the parallel component provoke a mechanical erosion. This action was greater than the normal component (Table would generate high temperatures and pres- 1). Generally, for all knives the normal cutting sures in the cutting area and on the tool edge force (F,,) increased as the planing length in- (Stewart 1989). For the jointed knives, the creased. The increase in normal force was contact area between the clearance face and greater between 0 and 500 m of planing. After- the new wood surface depended on the jointed wards, normal cutting forces increased at a land width, due to the zero clearance falling slower rate up to 3,000 m of planing. The par- to negative during the cut. This response was allel cutting forces for jointed knives decreased not seen in the unjointed sharpened knife. As as the planing length increased. For the unjoint- jointed land width increased, the magnitude of ed knife, the parallel cutting forces showed no Ifc ,i?~iirtlc:IIIZII ~lrMorrrcr- EFFECTS OF JOINTING AND WEAR ON PLANED SLJRFACE QUALITY OF RED OAK 545

apparent change as the planing length increased. The normal force component was therefore more sensitive to changes in wear compared to the parallel force component. The normal force increased 47% over 0 to 3,000 m of planing, while the parallel force decreased 7% (jointed lands pooled). Similar results were reported by Stewart (199 l), Huang (1 994), and HernAndez and Rojas (2002) while studying the effect of the cutting length on the forces in wood machining processes. The increase in the

\w' normal force due to wear should be greater in peripheral planing since the negative clearance Fig. 2. Normal orthogonal cutting force from knives produced is greater. as a function of the planing length for three levels ofjoint- The normal (F,) and parallel cutting forces ed land uidth (J.L.). (F,) tended to increase with increasing jointed land width (Table I). The normal tool force

Fig. 3. Maximum damage noted during planing northern red oak wood with a knife having 1.2-mm jointed land width and 3000 m of planing. 546 WOOL) AND FIBER SCIENCE, OCTOBER 2002. V. 34(4) was also more sensitive to jointing than the parallel tool force. The normal force increased 85% from 0 to 1.2 mm of jointed land width, while the parallel force increased 12% (planing lengths pooled). In general, the normal cutting force increased sharply from 0 to 0.4 mm of jointed land width (Fig. 2). The rate of increase of this force decreased slightly at 1.2- mm jointed land width, even though the edge recession continued to increase (Table I). This was similar to the effect of jointing on orthogonal cutting forces reported by Hernandez and Rojas (2002). As with wear, given that the clearance angle increases negatively, then in- Fig. 4. Maximum depth of damage during planing northern red oak wood as a function of planing length and creases in the normal force due to jointing jointed land width (J.L.) should be greater in peripheral planing. This analysis indicates that normal cutting forces are a good index of tool wear and surface damage was more severe as the jointed jointing. This index can be used to evaluate land width and wear increased. tools and wood species in other jointing sit- The freshly sharpened unjointed knife pro- uations. duced surfaces virtually free of damage (Fig. 5). However, the damage increased sharply up SEM ~znalysisof transverse glueline ,fuc.es to 500 m of planing and then more slowly from 500 to 3,000 m. The damage was 0.05 Crushed and collapsed cells were observed mm and 0.08 mm thick after 500 and 3,000 m near or at the glueline in transverse sections of planing, respectively. of the SEM samples. However, the type and When the two jointed knives were sharp- severity of surface damage were different ened, surfaces were little damaged (Fig. 6) but from those earlier reported for planed surfaces as jointed land width increased, damage in- of sugar maple (Hernandez and Rojas 2002). creased up to 0.06 mm deep at 1.2 mm of In fact, damage to red oak surfaces was less, jointed land width (Fig. 4). The damage in- with no layer of severely crushed cells. In this creased at a slower rate for both the jointed \tudy, damaged cells were seen mostly in thin- knives as planing length increased (Fig. 4). walled longitudinal parenchyma, vasicentric The damaged area varied between 0.07 mm tracheids, and vessel tissues, while thick- and 0.10 mm thick at 500 m of planing, and walled fibers were affected only occasionally increased to a range between 0.10 and 0.25 (Fig. 3). Differences in damage due to cell- mm at 3,000 m of planing (Fig. 4). Typical wall thickness have been reported previously damages under these cutting conditions are (White and Green 1980; Caster et al. 1985; shown in Figs. 3 and 7. Zink-Sharp et al. 1999). River and Miniutti The damage also tended to increase for all (1975) found almost no damage on the sur- planing lengths as the jointed land increased faces of red oak latewood and only moderate from 0- to 1.2-mm width (Fig. 4). This result damage in earlywood after cutting with saw, could be explained by the increase in the nor- planer, and jointer machines. We found that mal peripheral forces when planing with joint- the depth of the most severely crushed regions ed knives. A similar behavior was reported for in four typical specimens for each planing sugar maple wood by Hernandez and Rojas condition was a function of the planing length (2002). and jointed land width (Fig. 4). In general, the As expected, the roughness of planed sur- Ilr/~irr~i(l(,:(i~ul tlc hlortttr-EFIZECTS OF JOINTING AND WEAR ON PLANED SURFACE QUALITY OF RED OAK 547

Fig. 5. No damage vis~bleduring planing northern red oak wood with a knife without jointing nor wear. faces was directly related to the wear profiles sociated with the magnitude of normal cutting of the unjointed and jointed knives, and the forces. We confirmed that there was a statisti- surface roughness increased significantly as cally significant linear correlation between the planing length increased. Surface roughness normal orthogonal cutting force and the depth was not affected by the differences in jointed of the damage produced by planing. Even when land width within each planing length (not all planing lengths and jointed lands were shown). The increase in roughness was great- pooled, the normal tool force effect accounted est between 0 and 1,000 m of planing and in- for 58% of the total variation in the depth of creased at a slower rate up to 3,000 m of plan- the damage. This force and the surface rough- ing. Glueline thickness was also strongly af- ness were also significantly correlated. On the fected by the surface roughness. Peaks in other hand, no significant correlations existed rough surfaces prevented uniform contact be- between the parallel orthogonal force and the tween faces during clamping, resulting in two surface quality variables. thicker gluelines. Jokerst and Stewart (1976) as well as Stew- Gluing shear strength art and Crist (1982) suggested that the severity Average values for the apparent gluing of damage to the surface and subsurface is as- shear strength and the percent wood failure for 548 WOO11 AND FlBEK SCIENCE, OCTOBER 2002, V 344)

Fig. 6. Littlc damage noted during planing northern red oak wood with a sharpened knife having 1.2 mm of jointed land widlli. the four planing lengths, three jointed land tively (jointed lands pooled). The trend in per- widths, and the two conditioning treatments cent wood failure was similar. The moisture are given in Table 2. Means were compared saturation and subsequent drying cycles in the with the least-squares means statement from wood may have produced internal stresses re- the SAS General Linear Models procedure at lated to swelling and shrinkage. These stresses 95% confidence levels (SAS Institute 1988). would reduce the gluing shear strength and Under constant hygrothermal conditions, the wood failure by increasing the number of mi- gluing performance of red oak wood was not crofailures, particularly in the cell walls at the significantly affected by the different jointing surface and subsurface of wood. As a result, and wear processes. In contrast, the gluing the effects of wear on gluing performance shear strength and percent wood failure de- were evident after the accelerated aging treat- creased after the accelerated aging treatment. ment. Several authors have reported reduced The decrease in gluing shear strength was least gluing performance in wood after accelerated pronounced as the planing length increased, aging (Jokerst and Stewart 1976; Murmanis et and after aging, on average, gluing strength al. 1983; Caster et al. 1985; Hernandez 1994; was SO%, 3796, 32%, and 26% lower for 0, Hernandez and Naderi 2001). The aging treat- 500, 1,000. and 3,000 m of planing, respec- ment results are probably more representative If( III(~,I~~:

Fig. 7. Typical damagc occurred during planing northern red oak with a knife having 0.4-~nmjointed land width and 500 rn of planing. of the gluing performance of wood under nor- wear appears to be associated to the surface rnal working conditions. roughness. We found a statistically significant Aftcr the severe soak-dry treatment, the glu- polynomial correlation between the gluing ing shear strength significantly increased as shear strength and the surface roughness. wear progressed. The influence of wear was When all planing lengths and jointed lands different for red oak, a ring-porous wood, were pooled, the variation in roughness ac- compared to sugar maple, a diffuse-porous counted for 84.5% of the total variation in the wood (Hernhndez and Rojas 2002). Between gluing shear strength (Fig. 9). The actual wood 0 and 500 m oT planing. the gluing shear surface available for gluing was greater for strength greatly increased for all three jointed rougher surfaces, and the glueline was thicker land widths (Fig. 8). Between 500 and 3,000 since the adhesive filled the troughs in the sur- rn of planing, the gluing shear strength contin- face caused by the worn knives. Related to ued to increase but at a slower rate. The over- gluing shear strength, the percent wood failure all increase in shear strength due to wear was was also highly correlated with surface rough- about 39% (jointed lands pooled). In general, ness. Surface roughness variation accounted the trend in percent wood failure was similar. for 90.1% of the total variation in percent The increase in gluing performance due to wood failure (all cutting conditions pooled). 550 WOOID AND FIBER SCIENCE, OCTOBER 2002, V 34(4)

TABLE2. Averclge fil~rirzgshecrr strength (SS) and percent (fwood failure (WF)fi~rfi~urplaning lengths, three jointed Icrricl uidt11.r. trnd tvvo conditioning tremtment.c..

Con\tant cond~t~ons' After condir~nning' Pliln~ng Jo~ntcilIilnil length (~nj (mm) SS (MPd) WF (%) SS (MPa) WF (%I

I Mean, of twcnly rcpl~cant\. ' Standard error of Ihr mean In p:trenthe\r\ ' Mciln, w~lhan;I colulrln rr,llowctl hy the 5amc Icllcr are not \~pntfrcentlydlffcrcnt at the 5V pn,hahillty level. Upprrca\r lelter\ are tor jointed land cnrnpdr~wrand lowcrcaie letter\ are k)r planing Irngth compan\on. for each planing length. and lolntcd land \eparalely

The increase in roughness may have enlarged may have increased as planing length in- the actual surface available for adhesive pen- creased. The permeability in the transverse di- etration and for mechanical linkages at the rection of wood appears to be increased by the wood-glue interface, resulting in a more resis- wear process, with resulting higher adhesive tant glueline. However, optimal gluing perfor- penetration. Compression rolling treatment in- mance of red oak may have already been creases permeability in several woods (Cech reached (Figs. 8 and 9). Higher planing 1971 ; Giinzerodt et al. 1988). Cutting forces lengths and surface roughness could lead to generated during planing, particularly the nor- declining gluing performance. mal cutting force, increased as the planing On the other hand, adhesion penetration length increased (Fig. 2). Permeability of red

*0.4 rnm

-1.2 mm

0 500 1000 1500 2000 2500 3000 3500 81"""""""'""""'1 Planing length (m) 7 17 27 37 47 57 Rmax brn) Fig. X. Gluing \hear strength as a function of planing length for thrce levels of jointed land for samples that had Fig. 9. Variation in gluing shear strength as a function undergone an accclerated aging treatment. of surface roughness. Hrttl(itldc.- (ittd rlr, Moritrr-EFFECTS OF JOINTING AND WEAR ON PLANED SURFACE OUALITY OF RED OAK 55 1

creases the normal cutting force, could reduce the damage produced by knives. However, the probable increase in wood permeability due to the normal cutting force could be reduced. The method suggested by Stewart (1977) predicts that cutting forces that result from sharpened unjointed knives with a rake angle of 27 degrees could be optimal for producing a satis- factory wood surface. Experiments at different rake angles are needed to verify this predic- tion. Other factors that need to be considered

+3000 m include the feed rate per knife, diameter of the 7 cutting circle, and cutting depth. Given this 0 0.4 0 8 1.2 lack of information, general recommendations Jo~ntedland (mm) could be misleading and further investigations Fig. 10. Gluing shear strength as a function of jointed are required. land width for four levels of planing length for samples that hat1 undergone an accelerated aging treatment. CONCLUSIONS

oak, especially latewood, may have been in- The peripheral planing with sharpened un- creased by an analogous mechanism with glue jointed knives generated surfaces and subsur- penetrating via paths opened up in the late- faces of red oak wood with little damage. wood to reach sound cells where mechanical However, slight damage increased with in- attachments were available. creasing jointed land width and planing Therefore, under the given cutting condi- length. Damage was mostly in thin-walled lon- tions in this study, gluing properties of red oak gitudinal parenchyma cells, vasicentric tra- were enhanced by a surface with 40 pm R,,, cheids, and vessel cells, while thick-walled fi- roughness (Fig. 9) resulting from a 1.2-mm bers were relatively undamaged. The maxi- jointed land width knife, 0.33-mm rake face mum depth of this damage was positively cor- recession and 0.44-mm clearance face reces- related with the magnitude of the normal sion (Table I). Such increases in gluing per- cutting force. Knife wear also increased as the formance could be expected to progress up to jointed land and planing length increased, and an optimal planing length, after which the ef- this process was similar on both the rake and fect of knife wear would become negative. clearance faces of knives. The normal cutting The effect of jointing on the gluing shear force increased with wear and jointing. Within strength measured after accelerated aging is the wear and jointed land width levels studied, shown in Fig. 10. In general, there was no gluability was more sensitive to surface rough- significant variation in gluing shear strength ness than to cellular damage. The increase in among the different jointed land widths. The surface roughness from increased planing trend in percent wood failure was similar. length, caused a corresponding increase in the Therefore, the increase of the jointed land up gluing shear strength and percent wood failure to 1.2-mm width for planing had no effect on of glued specimens. The effects on these prop- gluing performance in red oak wood. erties were evident in specimens having un- In this study we used a rake angle of 20 dergone an accelerated aging treatment. In degrees, which is common in the wood indus- contrast, no significant variation in gluing per- try. The normal cutting force was relatively formance of red oak wood existed among the important even for sharpened unjointed three jointed land widths studied. The results knives. Increasing the rake angle, which de- suggest a jointed land of 1.2 mm as a maxi- 552 WOOL) AND FIBER SCIENCE. OCTOBER 2002. V. 34(4)

Inurn allowable width for planing this wood. HOADLEY,R. B. 2000. Understanding wood: A craftsman's Also, a surface roughness of 40 pm R,,,, pro- guide to wood technology. The Taunton Press, New- town, CT. 280 pp. duced by knives with 332 Frn of rake face HUANG,Y.-S. 1994. Cutting force components in orthog- recession and 438 pm of clearance face reces- onal cutting parallel to the grain (90-0) 11. Effects of sion, could be suggested as maximum limit for feed lengths. Mokuzai Gakkaishi 40(10): 1059-1066. positively affecting the planed surface quality JOKERST.R. W.. AUD H. A. STEWART.1976. Knife- versus of red oak for gluing. abrasive-planet1 wood: Quality of adhesive bonds. Wood Fiber 8(2):1071 13. KING,B.. AND R. 0. FOSCHI.1969. Crossed-ring dyna- ACKNOWLEDGMENTS mometer for direct force resolution into three orthogo- The authors thank Alain Clloutier, Robert nal components. Int. J. Mach. Tool Des. Res. 9:345- Beauregard, Rick Lemaster, Bryan River, and 356. KIRBACH,E., ANII T. BONAC.1982. Dulling of sawteeth Harold Stewart, for valuable assistance and tipped with a stellite and two cobalt-cemented tungsten Forintek Canada Corp. for technical support carbides. Forest Prod. J. 32(9):42-45. during planing tests. This research was sup- KOCH. P. 1985. Utilization of hardwoods growing on ported by the Minister of Natural Resources southern pine sites. Vol. 11. Processing Agriculture of Quebec, and the Natural Sciences and En- Handbook No. 605, USDA. Forest Service, Washing- ton, DC. 3,7 10 pp. gineering Research Council of Canada. KRIL~V,A. 1986. Corrosion and wear of sawblade steels. Wood Sci. Technol. 20:361-368. REFERENCES MCKENZIE,W. M. 1960. Fundamental aspects of the wood A~IIRICA~SOCI~ IY ~OK T~.STING AND MATERIALS(ASTM). cutting process. Forest Prod. J. 10(9):447-456. 1987. Standard methods for conducting machining tests MLJRMANIS,L., B. H. RIVER,AND H. A. STEWART.1983. of wood and wood-base materials. ASTM D 1666. Phil- Microscopy of abrasive-planed and knife-planed surfac- adelphia. PA. es in wood-adhesive bonds. Wood Fiber Sci. 15(2):102- (ASTM). 1994. Standard test method Ihr strength 115. properties of adhesive bonds in shear by compression RIVER,B. H.. AND V. I? MINILITTI.1975. Surface damage loading. ASTM D 905. Philadelphia, PA. before gluing-weak joints. Wood and Wood Prod. 80(7): CASTFR.D.. N. KUI.S(.HA.AND G. LEICK.1985. Gluability 35-36, 38. of sanded lurnben Forest Prod. J. 35(4):45-52. SAS INSTITUTL.1988. SASlSTAT User's Guide, Release CECH,M. Y. 197 1. Dynamic transverse compression treat- 6.03 Ed. SAS Institute. Inc., Cary, NC. ment to improve drying bchavior of yellow birch. Forest STEWART,H. A. 1977. Optimum rake angle related to se- Prod. J. 2 1(1):41-50. lected strength properties of wood. Forest Prod. J. 27(1): DLINSMOR~.E L*. 1965. The technique of woodworking 51-53. machinery. MacDonald, London. IJK. 173 pp. . 1989. Feasible high-temperature phenomena in GLINL~KOIIT.H.. J. C. E WALKER,AND K. WIIYBK~W.1988. tool wear from wood machining. Forest Prod. J. 39(3): Compression rolling of Sitka spruce and Douglas-tir. 25-28. Forcst Prod. J. 38(2):16-1 8. . 199 1. A comparison of tool materials, coatings, Ht~x~~1)t.z.R. E. 1094. Effect of two wood surface meth- and treatments related to tool wear during wood ma- ods on the gluing properties of sugar maple and white chining. Forest Prod. J. 41(9):61-64. \pruce. Forest Prod. J. 44(7/8):63-66. , AND J. B. CRIST.1982. SEM examination of sub- , AND N. NADERI.2001. Effect of knife jointing on surface damage of wood after abrasive and knife plan- the gluing properties of wood. Wood Fiber Sci. 33(2): ing. Wood Sci. 14(3):106-109. 292-30 1. WHITE.M. S., AN[) D. W. GREEN.1980. Effect of substrate . Aun G. Ror~s.2002. Effects of knife jointing and on the fracture toughness of wood-adhesive bonds. wear on the planed surface quality of sugar maple Wood Sci. 12(3):149-153. wood. Wood Fiher Sci. 34(2):293-305. ZINK-SHARP,A,, J. S. STELMOKAS,AND H. Gu. 1999. Ef- HILI.IS,W. E.. ANr) MCKENZIE,W. M. 1964. Chemical at- fects of wood anatomy on the mechanical behavior of tack a\ a factor in the wear of woodworking cutters. single-bolted connections. Wood Fiber Sci. 3 1 (3):249- Forest Prod. J. l4(7):310-3 12. 263.